DAVIDM. FRISCH AND DONALD w. VISSER
175G
in the proteins. The possibility that contaminating traces of formaldehyde in the radioactive formic acid were bound to the protein was ruled out by the finding of negative results on formaldehyde estimation.16 Since all radioactivity in the 25% sulfuric acid hydrolysate of the radioactive protein sample were recovered in the distillate in alkaline solution, the protein bound radioactive material must be formic acid. The amounts of radioactive carbon bound by various proteins seemed to be in parallel with their total hydroxyamino acid content (serine -I- threonine) as summarized in Table 111. Therefore the main reaction of formic acid with proteins must be 0-formylation of their hydroxyarnino acid residues. 0-Formylation of the tyrosine residue appears not to occur in anhydrous formic acid, since the paper chromatographic behavior and ultraviolet absorption spectra in both acid and alkali solution of the formic acid treated Nacetyl-L-tyrosine were the same as for the untreated compound. Since serine did not give 0-formylserine by reaction with anhydrous formic acid, 0-formylation of the hydroxyl group of serine (and threonine) probably takes place with the assumed intermediate oxazoline or hydroxyoxazoline compound Owing to the great differences in the stability of the 0-formyl groups of proteins observed in the present experiments, i t was not possible to esti-
Vol. 81
-NH-CH-C&NH-CH-C+
I
R
r
-NH-CH-C-N-CH--CO-
R
or
I
*CHI
I
HO-CH2
OH
I -NH-CH-C-NH-CII-COI I I
3.
R
0 -
1
CHI
HCOOH
-NH-CH-CO-NH-CH-CO-
I
R
I
HCO--O---CHa
mate the exact number of formyl groups introduced into the protein by the reaction between formic acid and protein. Formic acid may nevertheless prove to be useful as a specific reagent for the hydroxyl group of serine (and threonine) residues in protein, since i t appears to be milder and possibly more specific than concentrated sulfuric acid.22 Acknowledgments.-The author is grateful to Dr. H. Fraenkel-Conrat for suggestions and discussion throughout this investigation and to Mr. €1. Shichi for technical assistance.
(26) H. Fraenkel-Conrat, M. Cooper and H. S. Olcott, THIS JOURNAL, 87, 950 (1045). BERKELEY, CAL.
[CONTRIBUTION
FROM THE
DEPARTMENT OF BIOCHEMISTRY AND NUTRITION, SCHOOL OF MEDICINE, ~JNIVERSITYOF SOUTHERN CALIFORNIA]
5-Bromodeoxycytidine and 5-Chlorodeoxycytidinel
w. VISSER
B Y DAVIDM. FRISCH AND DONALD RECEIVEDDECEMBER 11, 1958
Two new deoxynucleoside derivatives, 5-chlorodeoxycytidine and 5-bromodeoxycytidine, have been prepared by a photocatalytic process. The inhibitory effect of these derivatives on the turbidity and viability of a thymine requiring mutant of Escherichia coli and a pyrimidine-requiring mutant of Neurospora are described.
In addition to the possible usefulness of these deIntrohction Several 5-halopyrimidines, their ribosides and de- rivatives in studies of nucleic acid metabolism, oxyribosides, have interesting inhibitory effects genetics and as chemotherapeutic agents, the new upon growth of bacteria, bacteriophage, molds and compounds are potential replacements for 5tumors. Some of these antimetabolites inhibit methyldeoxycytidine found in some plant and thymine synthesis, others prevent thymidine utili- animal DNA and 5-hydroxymethyldeoxycytidine zation, and depending to some extent on the size of in the coliphage DNA of the even varieties. the the abnormal pyrimidine may be Experimental incorporated into either RNA or DNA. Shive and 5-Bromodeoxycytidine.-Deoxycytidine (2.15 g., 0.0098 Skinner' have summarized these activities in a rewas dissolved in 155 ml. of anhydrous acetic acid, cent review. The unusual activities of the above mole), 110 ml. of anhydrous pyridine was added and the solution derivatives suggested the synthesis of two new de- chilled t o -10'. The cold solution was exposed to ultrarivatives, 5-chlorodeoxycytidine and 5-bromo- violet light at a distance of 9-10 cm. and 1.53 g. (0.0097 deoxycytidine, by a photocatalytic method similar mole) of bromine in 10 ml. of cold anhydrous carbon tetrawas added with constant stirring. The ultraviolet to that reported for the synthesis of 5-halo~ytidine.~chloride source was a Lamp Projector, model R3LS, manufactured (1) This investigation wns supported by the National Institutes of Health, Grant No. C2972. (2) 5. Zamenhof, B. Reiner, R. De Giovanni and K. Rich, J . B i d . Chem., 919, 166 (1966). (3) D. B. Dunn and J. B. Smith, Nature, 174, 306 (1964). (4) Shlve and Skinner in Ann. Rcr. Biochcm.. 97, 843 (1958). (6) T . K. Fukuhara and D. W. Visser, Tar8 JOURNAL, 77, 2393 (1955).
by the Keese Engineering Co. of Hollywood, Calif. The ambercolored solution was removed from the ultraviolet light and allowed t o stand a t room temperature overnight. The solvents were removed at temperatures not exceeding 40" using reduced pressure until a thick sirup was obtained. This material was dissolved in 30 ml. of absolute methanol and again reduced to a sirup. This process was repeated five more times or until a semi-crystalline white solid was
April
5,
0'
0
1757
5-RROMO- AND 5-CIILORODEOXYCYTIDlNC
1959
I
1
1 2 TIME (HOURS),
i
0
of 5-chlorodeoxycytidine on turbidity of Escherichia coli 15T- cells, plotted on logarithmic scale. Each curve represents a different conceutration of inhibitor: 0, control; 0 , 3.3 X 10-6 M; 0,1.6 X lo-' M ; m, 3.3 X 10-4 M . Fig. 1.-Effect
TIME
i
3
(HOURS),
Fig. 2.-The effect of 5-chlorodeoxycytidine on viability of Escherichia coli 15T- cells, plotted on a logarithmic scale. Each curve represents a different concentration of inhibitor: 0, control; 0 , 3 . 3 X 10-6 M ; 0,1.65 X lo-' M ; m, 3.3 X 10-4
M.
added as the growth requirement. As observed with the obtained. To this substance was added 200 ml. of cold ab- halo derivatives of uridine and cytidine,' the chloro derivasolute methanol solution containing 543% anhydrous am- tive is more effective than the bromo derivative and uridine monia. The mixture was shaken until the solid dissolved is a more effective reversing agent than cytidine. In the and then allowed t o stand at room temperature for three presence of sufficient cytidine to satisfy the pyrimidine redays. Methanol and ammonia were removed at room tem- quirement of the mutant and a concentration of chloro- or perature and reduced pressure. The amber-colored sirupy bromodeoxycytidine which inhibits growth completely, dematerial remaining was dissolved in a minimum amount of oxyuridine or deoxycytidine reverses the inhibition, whereas water and passed through an Amberlite IR-120 resin (chlo- thymidine is ineffective. Neither of the derivatives had ride form) column (2.2 X 25 cm.). The column was an apparent effect on the growth of wild type Neurospora. washed thoroughly with water to remove deaminated mateEscherichia coli cells (thymine-requiring mutant 15T-) rial. Ammonium hydroxide (0.4 N ) was passed through were grown aerobically in the synthetic medium of Cohen the column and the eluate, containing 5-bromodeoxycyti- and Abrogast' supplemented with 1 y/ml. of thymine. dine, was lyophilized leaving 2.8 g. of an amorphous, tan When about 7070 of maximal growth was reached, aliquots material. This material was dissolved in a minimum vol- of the culture were transferred to fresh medium so that the ume of hot absolute methanol, treated with Norite and fil- final cell count was 7-9 X 10' cells/ml. The fresh medium tered. Ethyl acetate was added t o the filtrate until the contained thymine to give a final concentration of 1 y/ml. solution became cloudy. The solution was then allowed and varying amounts of inhibitor to give final molar cont o remain uncovered at room temperature until crystalliza- centrations of 3.3 X 10-6, 1.65 X 10-4 and 3.3 X lo-'. tion occurred. Successive crystallizations yielded 1.9 g. Samples were taken at 0, 1, 2 and 3-hour intervals. Tur(620J0)of 6brornodeoxycytidine, m.p. 175-179', Xi,".: 300 bidity was measured in a KlettSummerson colorimeter mp (e 9,600). Anal. Calcd. for GHlnOdNIBr: C, 35.31; using a 420 mp filter. Viability was determined by spreadH,3.95; N , 13.73. .Found: C.35.39; H,3.89; N, 13.76. ing small aliquots of diluted cultures on broth agar plates.' 5-Chlorodeoxycytidme was prepared in an analogous man- The highest concentration of 5-chlorodeoxycytidine proner to the Bbromodeoxycytidine, yield 38'70, m.p. 184- duced a 40% decrease in turbidity as compared t o the con1m0, 295 mp (e 9,980). Anal. Calcd. for C ~ H I I O trol ~ a t 3 hours; see Fig. 1. The viability dropped rapidly NSCl: C, 41.31; H, 4.62; N, 16.06. Found: C, 41.36; during the 3-hour period until less than 1% of the cells reH, 4.71; N, 16.50. mained viable; see Fig. 2. The curves for similar conccnMicrobiological.-The deoxycytidine derivatives were trations of 5-bromodeoxycytidine werc substantially the tested for their effect on the growth of Neurospora using same as for the 5-chloro derivative. Microscopic exammethods described previously.0 Both conipounds com- ination showed that the inhibited cells were about 5 times petitively inhibit growth of the pyrimidine-requiring mu- the size of the controls. tant, Neurospora 1298, when either uridine or cytidine are (6) M. Roberta md D. W. Visser, J . B i d . Chcm., 194, 695 (1952).
(7) S. S. Cohen and R. Abrogsst. J . E x g f l . hfed.. 01, 619 (1850). (8) S. S. Cohen. J . Bid. Chcm., 168, 511 (1947).
ERICHBAERAND DMYTRO BUCHNEA
1758
The brornodeoxycytidine derivative produces a high frequency of plaque-type mutations in T-2 bacteriophagc.e Screening results at Sloan-Kettering Institutelo have shown that 5-chlorodeoxycytidine has an inhibitory effect on tumor growth (Ca 755) in rodents, whereas the 5-bromo derivative is inactive.
Discussion All of the 5-chloro- and 5-bromopyriniidine ribosides and deoxyribosides previously tested competitively inhibit utilization of uridine by the pyrimidine-requiring mutant, Neurospora 1208. The competitive nature of these inhibitions,6,6and the fact that the wild-type Neurospora is not inhibited by comparatively high concentrations of these compounds indicate that the primary site of inhibition in the nutant probably involves the conversion of uridine or cytidine to the nucleotide by nucleoside kinase. In E. coli, 15T- the deoxycytidine derivatives inhibit growth and decrease viability in a manner similar to that described for 5-bromodeoxyuridine.l 1 (9) J. Gregory, private communication. (IO) D a t a of D. Clarke, K.Sugiura a n d C. C. Stock. (11) S. S. Cohen a n d H. D. Barner, J . Bocferiol., 71, 588 (1956).
VOl. 81
Thus, it is possible that the deoxycytidine derivative exerts its effect by a prior deamination to bromodeoxyuridine and subsequent incorporation into DNA in place of thymidine.l2 However, other interesting changes in DNA structure may result from the presence of the deoxycytidine derivative, such as incorporation into DNA in lieu of deoxycytidine, or formation of an abnormal base as with 5aminouracil. l 3 These possibilities are under investigation. The high frequency of plaque-type mutations in T-2 bacteriophage produced by bromodeoxycytidine also suggests the possibility that the analog may be incorporated into phage DNA. either as the halogenated deoxyuridine, in place of thymidine, if deamination takes place, or as the halogenated deoxycytidine in place of viral 5-hydroxymethyldeoxycytidine. If the latter occurs, i t would be an interesting model for chemotherapy application. (12) S. S. Cohen a n d H. D. Barner, J. B i d . Chem., 396, 631 (1957). (13) D. B. D u n n a n d J. B. S m i t h , N o f u r e , lT6, 336 (1955).
Los ANGELES,CALIF.
(CONTRIBUTION FROM THE SUBDEPARTMENT OF SYNTHETIC CHEMISTRY IN RELATION TO MEDICAL RESEARCH, BANTING AND BEST DEPARTMENT OF MEDICAL RESEARCH, UNIVERSITY OF TORONTO]
Synthesis of L-a-(Dioleoy1)-cephalin ; with a Comment on the Stereochemical Designation of Glycerolphosphatides’ BY ERICHBAERAND DMYTRO BUCHNEA RECEIVED JUNE 30, 1958 A synthesis of L-a-(dioleoy1)-cephalin is reported. The unsaturated cephalin is obtained by phosphorylation of D-a,Bdiolein with phosphorus oxychloride and quinoline, immediate esterification of the resulting L-a-dioleoylglycerylphosphoryl dichloride with 2’-hydroxyethylphthalimide in the presence of pyridine, and removal of the protective phthaloyl group by hydrazinolysis. The configurational and structural purity of L-a-(dioleoy1)-cephalin was confirmed by catalytic reduction t o L-a-( distearoy1)-cephalin, and comparison of its specific rotation with that of authentic material. The infrared spectrum of L-a-(dioleoy1)-cephalin, and the solubility of the substance in varinus solvents are reported.
The synthesis of structurally and configurationally pure glycerolphosphatides containing unsaturated fatty acid residues has lagged behind that of the saturated analogs, although unsaturated phosphatides are known to occur more widely in nature. I t is only in the last few years that improvements in the techniques for the separation of lipid mixtures have made possible the synthesis of such unsaturated compounds via less protected intermediates than those used in the saturated series. Two years ago we reported the synthesis of L-a-(dioleoy1)-lecithin.2 The syntheses of the nitrogen-free phosphatides, dioleoyl L-a-glycerylphosphoric acid, 3a tetraoleoylbis- (L-a-glyceryl) -phosphoric acidaa and (dioleoyl-L-a-glycerylphosphory1)-L-a-glycerolahare reported by us in two recent papers. The purpose of the preseiit paper is to describe the synthesis of ~-a-(tlioleoy1)-crphali11. It is interesting to note that oleic acid appears to be the principal un(1) T h e synthesis of i.-~r.(dioleoyl)-ce~,Ilalin was reportcd in i i lectiire presented a t a “Syml)osium on Phosplioric Esters irnd llclatecl
Compounds” held by t h e Chcmical Society a t Camliridgc, England. April 9-12, 1957. (2) E. B a a , D. Buchnc,n und A. G. Newcornlir, Tlrrs IOWRNAI., 78, 232 (l95G). (3) (n) E. Baer a n d D. Buchnea, Arch. Biochem. a i d Rioghys., 78, 294 (1!458); (b) J . Biol. Chrm., 232, 895 (1958).
saturated fatty acid in naturally occurring cephalins.4J As mentioned in an earlier publication2 the procedures developed in this Laboratory for the synthesis of saturated a-phosphatides, employing phenylphosphoryl dichloride as phosphorylating agent, are not suitable for the synthesis of unsaturated phosphatides, since the removal of the phenyl group by catalytic hydrogenolysis would lead to a simultaneous reduction of the fatty acid double bonds. At the time i t was, however, felt desirable to retain the use of phenylphosphoryl dichloride for the synthesis of dioleoyllecithin and a method was devised which made this possible.2 The L-a-(dioleoyl) -lecithin was obtained via the following sequence of intermediates : D-acetoneglycerol --t acetone-L-a-glycerylphenylphosphorylchloride + acetone-L-cu-glycerylpherlylphospliorylethylene chlorohydrin + r,-cu-glycerylphosphoryl ethylene chlorohydrin -+ L-cu-ctioleoylglycerylpIiosphory1ethylene chlorohydrin --t L-cu-dioltoylglycerylphosphorylcholine. I t was hoped that 1,-a-(dioleoy1)-cephalin could be obtained by using the same procedure, J. Parnas. Biorhrvi 2 , 2 2 , -111 (1009). ( 5 ) C . G. MacArthur and L. V . Burton, T H 1 9 J O U R N A L , 38, 1375
(.4)
(1916).
